DESCRIPTION (provided by applicant: Very few commonly used neonatal surgical therapies have been evaluated to determine their safety. Premature infants are especially vulnerable to unrecognized treatment hazards. Inguinal hernia (IH) repair is the most commonly performed major operation in premature infants, is associated with a 30-40% morbidity rate, and it is unknown whether the operation should occur prior to neonatal intensive care unit (NICU) discharge or be delayed. Early IH repair possibly increases anesthesia and cardiorespiratory serious adverse events (SAEs), whereas delayed IH repair possibly increases hernia-related adverse events. A growing concern is the possibility that early general anesthesia exposure impairs neurodevelopment. The randomized clinical trial (RCT) proposed will be the first trial to evaluate the effect of timing of general anesthesia exposure on neurodevelopmental outcomes. Our objective is to conduct a 19 center RCT to determine whether early IH repair (prior to NICU discharge) or late IH repair (~5 months after NICU discharge) is the safer surgical approach for premature infants who are diagnosed with an inguinal hernia while in the NICU. Assuming a 60% randomization rate, which is supported from a feasibility pilot RCT, 651 infants will be randomly assigned to early or late IH repair over a 2.5-year period. The primary hypothesis is that late IH repair is safer as defined by a 10% absolute reduction in the proportion of infants with >1 SAE (relative risk = 0.66; number needed to harm of 10 for early repair) within 10 months after enrollment. The primary outcome, as adjudicated by a committee masked to treatment group assignment, is the proportion of infants with >1 SAE during the study period. Secondary outcome measures include the number of hospital days during the study period (hypothesis: median 3 days less with late IH repair) and, in selected centers (n=200 infants), neurodevelopmental testing with the Bayley Scales of Infant Development (BSID), 3rd edition, at 2 years corrected age (hypothesis: 0.4SD higher Cognitive Composite score with late IH repair). We will also measure non-routine events during the IH repair operation, which may underlie SAEs. The potential differences in treatment effect according to pre-specified patient characteristics (at enrollment) will be examined by frequentist and conservative Bayesian methods. The trial is utilizing the Pediatric Surgery Research Collaborative, a cooperative group committed to performing prospective clinical studies in pediatric surgery. Successful completion of the RCT proposed will stimulate future safety studies within neonatal surgery.